LONGEVITY

Risk: How important is longevity risk within life insurance and

pensions? How aware are pension funds about the risks within

their policies?

Thomas Streiff (TS): Pension funds and insurance companies are

well aware of the risks. What to do about the risk is the bigger

challenge. Managing longevity is extremely important and it is a

significant risk. If you just look at the annuity market place in the US

– which is where most of the annuity business is today – it is an

extremely large business and has been growing rapidly. In the first

half of this year, it experienced 10% growth over last year. The living

benefits in annuities are much more complicated than a typical

pension annuity, which is a straight life annuity or joint life annuity.

Christopher Murphy (CM): The actuarial profession has been

tasked with predicting mortality through the years, but history tells

us that it’s extremely difficult to get these assumptions right and

has highlighted the difficulty in dealing with this issue.

Risk: Are pension funds aware of some of the deficiencies of the

data on mortality?

Matthieu Robert (MR): We don’t have full transparency, but pension

funds have been slow to update their estimates of life expectancy. In

the UK, some companies have recently been forced to be more

conservative and increase the life expectancy of their policy holders

under IAS19 and the Pension Protection Act, however, there are still

some using aggressive assumptions. As for their awareness, we can

see from the end of the 1990s and the beginning of this decade that

it is only when long-dated interest rates dropped to historical low

levels that they realised they had problems with their guaranteed

annuity obligations and the potential losses they were facing. They

bought a lot of long-dated receiver swaptions to cover their risk. It is

interesting to see how quickly they went from no hedge to

executing huge hedging programmes against falling interest rates.

Nicolas Tabardel (NT): If we go back to before 2000, the typical

pension fund was heavily invested in equities, and equities were

going up so longevity was not too much of a concern to anyone.

After 2000, equities went down and rates fell, which increased

liabilities and people started to realise that there is a lot of risk here.

Since then, we have seen more and more liability-driven

investment (LDI) and, in a typical LDI approach, the manager of the

pension will try to quantify and hedge inflation risk, interest rate risk

and longevity risk. The first two are now quite easy to deal with, but

for longevity risk there are a lack of instruments, although there is

more and more will to deal with this.

Evan Guppy (EG): The reason we are seeing pension buyouts is that

people are really concerned about interest rate risk, inflation risk and

longevity risk. There is a premium they can pay in order to make those

Uncertainties about future mortality and life expectancy are proving a problem for pension

funds and insurance companies. They are increasingly looking to hedge longevity risk, but

the market is lacking instruments for it to hedge against. Last month, in conjunction with

Tullett Prebon, Risk hosted a roundtable to discuss the ways of assessing longevity risk and the

development of such a market

Longevity and mortality risk

The Panel

Barbara Blasel Global risk solutions, BNP Paribas

Benoit Chriqui Director, Head of inflation trading, Barclays Capital

Evan Guppy Manager, Inflation structuring, HSBC

Nadir Latif Business development, Member of Executive Committee,

Tullett Prebon

Christopher Murphy Managing director, Morgan Stanley

Yan Phoa Director, Derivatives trading, Lehman Brothers

Michaël Picot Director, Alternative derivatives trading, Calyon

Matthieu Robert Managing director, Head of inflation, Dresdner Kleinwort

Thomas Streiff Managing director, UBS

Nicolas Tabardel Vice-president, Inflation trading, Citi

Thomas Streiff

1107_Risk_longevity.indd 44 30/10/07 11:17:47

problems go away. The problem is that the cost of that risk transfer is

too great for many pension funds, which is why we need capital

markets solutions to bring in more participants to take on that risk.

Risk: Do you feel that the clients you are speaking with are

moving more towards the buyout solution or are people crying

out for more fi nancial derivatives to manage longevity?

Yan Phoa (YP): The emergence of the bulk buyout business was

driven by a big change in the regulatory and accounting regimes

for pension funds. The UK has probably gone furthest in valuing

and marking to market these liabilities through the corporate

balance sheet. Rating agencies have started to focus on them too.

For the pension fund trustees, there is much more personal liability

to ensure they are acting in the best interests of the pensioners. So,

overall, people are much more aware of risk. Pension trustees are

looking much more at whether they are adequately funded from a

personal liability perspective. What this has done is make

corporates realise that they don’t want to have this unknown

liability hanging over them. This, in turn, has driven the emergence

of this buyout market where people are willing to take on that risk

at a premium to where, historically, those liabilities had been

valued. Insurance companies have to price it for the uncertainty of

the future. From the pension funds’ perspective, as long as the

corporate sponsor is fi nancially sound, any mistakes in the

estimation of this risk can be rectifi ed by going back and changing

the contribution they are receiving. A pension fund will mark it

only to where it is reasonable, whereas an insurance company has

to mark it for future uncertainty and that naturally is done at a cost

as they will have to be more defensive in their estimates.

Risk: There has been a call from the media for pension funds to be

more transparent in their mortality assumptions. Is this possible?

Barbara Blasel (BB): It is possible, although the modelling and the

determining of the data is an issue. All the evidence points to a

systemic underestimate of future longevity improvement at all times

in the past. We somehow have to get there to determine longevity

and improvement in mortality in order to build up a market, but it

will take some time to fi nd standardised mechanisms to do so.

CM: If there is a deep and liquid market in longevity risk, this problem

would go away. We would no longer need to make the assumptions.

If you go back 10 years, pension funds needed to make infl ation

assumptions, now they don’t. The market decides it for them.

Risk: What role have reinsurers played so far? Is that the best

route to take?

TS: We are going to see an increasing role for reinsurers. We now

see reinsurers emerging as a potential solution for all of these risks

and taking them in a package. Most investment banks look at the

risk of infl ation and capital markets as part of the things that they

do, but they don’t necessarily want to deal with longevity risk or

behaviour risk that we are talking about in annuity, so they want

that to go somewhere else, whereas reinsurers could take on all of

the risk and then distribute what they don’t want.

Benoit Chriqui (BC): The advantage of reinsurers is that they can

be fl exible in terms of structures or formats, they can be quite

accommodative. But they are not going to be the driver of the

market. First, they are interested in risk diversifi cation, and longevity

risk is just one type of risk, even when you consider very diff erent

pension schemes, so they will be limited in terms of size of

transaction they are willing to do. Second, they are relative-value

players, so they are only going to be interested in the trade if they

can collect a premium to where they see fair value. The pensions

on the other hand might only be interested in hedging if it is close

to where they are currently valuing their liabilities, so might be

reluctant to trade at the premium demanded by reinsurers.

BB: Reinsurers have some kind of diversifi cation by not only

absorbing longevity risk but also mortality risk. It’s not a perfect hedge

but there is some degree of diversifi cation that probably allows them

to absorb more risk on the longevity side. But there is one constraint

on the reinsurance side, which is capacity. In recent years, there has

been a decrease in the number of reinsurance players.

Michaël Picot (MP): Giving all the risk to a handful of companies is

not the targeted solution. Reinsurers might have the capacity to

generate more returns than the pension funds, but they would

need to extract the longevity component and distribute it to the

market. This is precisely a role longevity derivatives can play.

Risk: There have been attempts to launch longevity securities;

BNP Paribas was involved in one of the bonds in late 2004. How

do you describe the attempts to get this off the ground so far?

BB: The feedback was very positive. That was the fi rst attempt to

hedge against longevity trends. It was probably a bit too early for the

market, therefore ultimately the bond hasn’t been issued. The reasons

for that are manifold. One was the introduction of the pension reform

at the same time, so pension funds were more focused on that. Also, a

supranational was the issuer, so yield was lower than some participants

were expecting. On the other hand, to provide a stable, long-dated

asset, it was important to have a high rating from the issuer.

Risk: What is the potential for longevity-linked securities?

NT: One issue with the bond format is that the buyer then takes on

the credit risk of the issuer. If you’re trying to hedge longevity risk

with a () bond, you introduce a new problem with the credit risk,

and that is at least as important as the longevity risk you were

concerned with in the fi rst place. Maybe a derivative format is better,

because the credit risk can be managed with a collateral agreement.

CM: That was the problem with the BNP Paribas structure in that it

wasn’t leveraged in any way. So, if you wanted to hedge out your

longevity risk, you would have to put 100% of your assets in this

bond. That is just not a feasibly solution. At Morgan Stanley, we

took a mandate from an insurer and worked on a similar type of

project for 12 months where we designed a bond that would pay a

SPONSORED ROUNDTABLESPONSORED ROUNDTABLE

Christopher Murphy

Nadir Latif

1107_Risk_longevity.indd 45 30/10/07 11:17:57

LONGEVITY

high coupon that would be reduced if longevity turned out to be

greater than current assumptions. The problem was that at least

half of the investors were not interested and 80% of the remaining

half was not allowed to invest in this product.

YP: Having a funded instrument is not going to be how the market

moves forward. The other problem with the bond is that, for it to

have some sort of liquidity, it needs to be large and a one-size-fits-

all solution. It is going to be hard to get enough people to feel

comfortable that the basis risk is acceptable.

NT: There is also the issue of supply. We know where the demand

will come from: pension funds. But who is going to supply the risk

that we need? There is a partial hedge in mortality risk, but there is

still a need in the tail of the distribution that we cannot source.

Risk: What are your views on the basis risk issue and the one-

size-fits-all argument?

TS: I’m optimistic about a future for an index approach but my

optimism is tempered by the issues raised. If I separate the market

into annuity and pension funds, the basis risk is probably more

manageable on the pension side. It’s still an issue but a bit more

manageable. On the annuity market, because those are voluntary

buyers and you have natural anti-selection with voluntary buyers,

the basis risk can be much bigger.

CM: Doesn’t that suggest that you just need another index?

TS: It may, in fact. If you can match an index that takes into account

your anti-selection then you’re in business. That is difficult to do

because what you’re trying is to get the buyer to be the seller and

that is where your challenge is.

YP: On the annuity side, it’s a much more concentrated market and

the competitive advantage of annuity providers is the information

they have. That data is proprietary and for them to say they are

going to contribute in-house data that is their competitive

advantage into an industry effort is going to be tough.

TS: You can get them to contribute the data such that they are

giving you meaningful information by virtue of what they are

buying and selling as opposed to what they collect in their own

database. That is hard to achieve but that is how you will end up

with an index that is matching what is actually happening in a

place where you have voluntary buyers.

EG: It is going to be quite difficult for a funded instrument to work

because the trustees and pension firms that are most aggressive in

trying to address their problems are the guys that are pursuing

interest rate and inflation strategies. For them to change their

approach to longevity and to reintegrate their asset and liability

management seems unlikely. But, if you’re trying to go down the

derivatives route, then sure, there is no reason why it couldn’t work.

MR: Concerning the index, if you want an interdealer derivatives

market on longevity risk to develop, you have to agree on one or

two indexes. It is most likely that we will have global indexes (i.e.,

on national population) trading. Pension funds will have the choice

to hedge their exposure with the global index and keep the basis

risk or to try to pass the risk relative to the subindex to which they

are exposed to a bank. With the first approach, I think it is possible

to work with the pension fund on historical analysis and

comparison between the global population and the particular

population of the subindex in order to come up with some

appropriate hedge ratio between the pension fund risk on the

subindex and the hedge on the global index.

Risk: There have been a number of attempts by banks that have

started developing indexes. Is there an obvious candidate as to

the type of index we could use? What are the obstacles?

BB: The idea should be to work on some kind of standardised index

because it’s the only way to allow for the development of a

derivatives market and to have transparent indexes that are tradable.

On the other hand, even setting up these indexes is already an issue

because forecasting longevity expectations for the next 30 years is

quite hard. We have to start with population mortality data because

this database has sufficient historical information. To set up forecasts

for the next 25–30 years, one would need an observation period of

30 years or even more. It would be great to have also subindexes

that cover certain groups of the population, such as specific

professional groups, and also cover regional differences within one

country. But, in the end, it comes down to availability of robust data.

Although longevity is seen as a global risk, risk sellers would

definitely want to have their particular geographic exposure

reflected, so one could possibly think of a handful of indexes for the

main markets from which companies can choose and design their

individual derivatives to reflect their specific exposure. This would, in

turn, become their individual exposure.

MP: Even population mortality rate is a debatable issue because it

is a difficult figure to produce, for example, due to infrequent

population censuses. Still, I expect to see two types of indexes. The

first one would be life expectancy index, just quoting life

expectancy at birth. The second approach is the life metrics

approach based on explicit qx ratio, which is much more ambitious

and therefore allows things like quoting the different life

expectancies at a given age. It also gives more room to hedge

longevity with mortality and a simple way to address the basis risk

with a ‘technical age’ approach (sliding on the curve to project one’s

specific population on the general population).

BC: On the derivative side, I agree it is impossible to have the market

develop with multiple indexes. There is not enough two-sided

interest to create many small markets, so you need to have everyone

focus on one specific index to help improve liquidity. The problem

on the pension side is that they will not view a central index as a

particularily good metric for their own longevity exposure. There

needs to be education to the pensions in terms of showing and

explaining the correlation between different indexes and how using

Matthieu Robert

Nicolas Tabardel

1107_Risk_longevity.indd 46 30/10/07 11:18:10

SPONSORED ROUNDTABLESPONSORED ROUNDTABLE

a central index would not be a bad hedge. But even then, it will still

be diffi cult, because although a central index would be a good

hedge, if it is used to value the liabilities, it could still give a very

diff erent value because of basis diff erence, and pensions might have

a hard time not valuing their liabilities at the same level as their

hedge. The other point is that there are lots of banks thinking about

indexes and how to construct them, but the fi rst diffi culty is data

collection. You also need to have a robust metric and have a

transparent way of collection that is unbiased. On that note, it would

help if the government got involved and stepped in to recommend

a metric to be used in valuing longevity liabilities, and then have an

offi ce that would be in charge of collecting the data.

YP: We need a transparent and impartial data collection method.

The government fi ts that role very well. You got to have the

commitment to produce the data on an ongoing basis. We’re

talking about transactions that are going to span the next 20–40

years. The market needs to feel comfortable that the data is going

to be provided in a consistent manner. It needs to be timely, you

need a long historical series so you can understand the behaviour

of the index in the past. There will be basis risk between how their

particular population behaves versus the general population, but

what you’re trying to hedge is the change in the trend in the future.

EG: I don’t think the government needs to get involved. There are

examples in the property market and indeed in credit derivatives

where it is quite possible to have an independent market-wide body

to collect the data. The key point is really more about independence.

If you want to use derivatives to draw more players into the market,

that means you want data that is supplied by someone who is visibly

independent, visibly reliable. If you’re saying you want life insurers and

pension funds to be trading with money managers and hedge funds

based on data that is collected by the insurers and the pension funds,

then it’s probably not going to be very appealing for the money

managers. The government is already collecting that for the widest

possible population set you can have in the UK. The other point is

how do you use that to construct a tradable index? I think people

would be more comfortable if, as a group, we come up with some

methodology and some index that would either be an independent

market-based index that could evolve out of something that was one

fi rm’s idea but is ultimately owned by the whole market.

NT: I see a parallel to the infl ation market. There is a basic trade-off

between liquidity and basis risk in trading one single index versus

multiple indexes. In the infl ation market in Europe, the market is

concentrated on European infl ation – that’s where the liquidity is. It’s

transparent and liquid. On the other hand, you have fi rms’ liabilities

that are not indexed to European infl ation but that use European

infl ation to hedge. They can use French infl ation, or Italian or Spanish

but it’s less liquid and is more expensive. It is less transparent. We are

always going to have this kind of trade-off . Whether we trade a single

longevity index or several, the question is: who is going to

warehouse the basis risk? Is it better to warehouse the basis risk in

banks or in the pension funds or clients? If the bank is required to

provide customised index for clients, then it will have to hedge with

a liquid index and warehouse the basis risk. If it has a cost, is it more

effi cient to have a cost in the bank or to have the cost at the client?

MR: There is no choice at the beginning but to go for a

standardised index. What is important is the transparency and the

independence in the building process of the index. I am quite

sceptical of any index that is built by a bank, even if it is relatively

transparent. I would typically prefer something built by a

government institute. For example, the Offi ce of National Statistics

(ONS) in the UK has already started collecting data on mortality

rates for the country. The ONS data is published on a annual basis.

The idea will be to use this set of data and have a panel of banks

agreeing on how to construct the index based on this set of data.

TS: Standardisation is a good idea. A certifi ed pension actuary

society and a society of actuaries on the insurance side could play a

role in gathering the data. They do a better job of gathering and

analysing the relevant data. In stable population countries, generally

mortality data is better than in countries where you have a lot of

turnover in the population through emigration or immigration.

General population data in the US is skewed because of the high

level of immigration. That is an example where we would never rely

on that data. We would rely on Society of Actuaries data or pension

actuary data if it’s on the pension side. That’s where we have

historically looked for standardised data. That is exactly what you are

doing in buyout situations: you are placing the longevity risk with an

entity that understands the data from a very diff erent perspective

than what you get in a generalised government index. I’m in favour

of a standardised index, but it’s not as big an answer as might

otherwise be viewed because of specifi c challenges of the data.

Nadir Latif: There has been a lot of discussion about indexes based

on government statistics. Would the market prefer an index

provider that is owned by several of the large banks and/or the

interdealer brokers? In terms of developing the market place,

getting a group of banks to agree to take joint ownership of a

company that is independent in constructing the methodology of

the index might be the quicker way to develop the market rather

than going down the route of a link with a government-based

entity. Of course the independent company could still base the

index on government-based data.

BC: If I were a pension fund, I would disagree with a group of banks

gathering the data to value my liabilities. If diff erent banks in the

group had diff erent interests, then this method would be fi ne, but

here banks are likely to always take the other side of the pensions.

The most important thing in this case is to get the pension funds

comfortable with the level of impartiality and that means not

owning the collection process or the creation of the index process.

EG: One of the advantages of having an independent fi rm collect

Evan Guppy

Yan Phoa

1107_Risk_longevity.indd 47 30/10/07 11:18:21

the data is that, as soon as the firm is up and running and

publishing the index, you then have the banks that have

committed themselves to supporting the index.

YP: One concern is continuity of such an index. It has to be a long-

term commitment. You have to have the commitment to see the

market succeed, at the same time, you have to make the

commitment to fund the index.

NT: As a client, I would be very wary of an index that is sponsored

directly by banks, whether by a single bank or a group of banks.

What happens if my bank doesn’t sponsor the index? Do I get the

data later, do other people have better information? That’s the

dangerous issue. There is also an issue of the continuity of such an

index. It’s hard to guarantee that as a private entity that is owned

by a group of banks. The only entity that can guarantee that over a

lengthy time horizon is the government.

CM: It needs to be based on government data, but there is no way

that the government is going to start publishing an index, so it will

have to be an independent index provider but maybe based on ONS

data. The involvement of actuaries in publishing that index would be

detrimental to the success of that index and there is widespread

scepticism about the value-added of actuaries in that process.

BC: The end-user should ultimately decide. There should be more

fact-gathering as to what type of central index would make sense

to somebody who has exposure to a subindex. Because the basis

risk is so important, there is a high likelihood that, if banks get

together and set up an index without properly consulting the end-

user, that this index will not be well received.

Risk: Is there any interest from clients about having a solution

built for them? Is there consensus over the methodology for a

standardised index?

BB: Clients are relatively open in terms of the types of solutions

they are looking for. In terms of indexes, market players need to get

comfortable with the modelling, and then transparency and

continuation are key in this context. As long as one can agree on a

single methodology and really continuously focus on this

methodology, then this is the driver for market participants getting

comfortable with the modelling and the analysis.

CM: Having a level of subjectivity in the creation of the index is

probably something that the pension and life insurance companies

would want to see because that is what they are comfortable with

and used to. But if you are talking about attracting new pools of

capital, like hedge funds, to this market, we will have extreme

scepticism about any potential tweaking to the underlying data

that is at all subjective.

YP: I think that is a fair point. There’s trying to get the best

information out of the data and trying to keep the data as close to

the raw form as possible. I would argue for trading that each

institution can do what it wants with the data in-house. What you

actually trade and settle the trades against, you may find that you

always want to revert to the raw data.

TS: And that’s how mortality tables are used now in the insurance

industry. They take a standard mortality table and adjust it for their

circumstances. So they will take what is, in essence, raw data and

put their own subjectivity on it based on how they want to price

that business. So I can see exactly the same thing happening here.

CM: To some extent, banks are indifferent as to how the index is

constructed. It matters more to the end-users who are going to

end up wearing the risk. And so, maybe the pension and insurance

industry should be discussing this topic as well and saying what

they want the index to look like.

Risk: How big do you actually see this market becoming? Are we

seeing any swaps and forwards transacted at this point in time,

or is it a little too early?

BC: There is some activity. But there are a couple of issues on which

we have touched that make it a little difficult to see how this market

is going to develop into something very large such as the inflation

market. There is significant end-user interest, and defined benefits

schemes are heavily exposed to longevity. But you have a very one-

sided business. One problem I see with this market developing is

that the end-users are currently using their own assumptions. If they

used the buyout assumptions, it would be more expensive for them.

If a market started to develop, where would that market start to

establish itself ? At the buyout assumption? Even more expensive

than the buyout level? And then, once you have an actual active

market, these pension funds to some extent will have to market-to-

market. Even if they claim their basis is different, it will become more

difficult to justify the assumptions they are using. To some extent,

the end-users who most need this market to hedge their liabilities

might also be reluctant to see it develop in a transparent way.

EG: There is a definite first-mover advantage for the pension

schemes that do something if a market gets up and running. Part of

the reason for going down a capital markets and derivatives route is

trying to get more players involved. The question is whether there

are sufficient investors interested in taking on some of that risk to

offset the £1 trillion in pension liabilities. That is the great unknown

at the moment and, until the market gets up and running, it is

difficult to see whether money managers, hedge funds or anybody

else are actually going to be willing and how much money they are

going to be willing to put up and at what price. One thing I would

hope is that, if you are going to have more people involved in that

space that can take on the mortality risk, then hopefully the market

will start to offer inside where buyouts are at the moment.

CM: This market may take a lot longer to develop than perhaps

we’re all hoping for because you can argue now that pension funds

potentially shouldn’t be marking to an FRS-17 mark, they should be

marking to buyout. Buyout prices include the longevity risk

associated – that is the only thing that’s left to hedge out. I don’t see

Barbara Blasel

Benoit Chriqui

LONGEVITY

1107_Risk_longevity.indd 48 30/10/07 11:18:32

SPONSORED ROUNDTABLE

any appetite at all for doing that. Clearly, legislation has gone a long

way to getting pension funds to address the risk embedded in their

liabilities and mark them to market, but the industry has been

kicked so much that there is no great appetite to create more stress

in the system. I’m sceptical as to how this market will develop. I’m

not so optimistic that it will come inside the buyout price. For this

market to take off, you need new pools of capital to come in and, if

money is made available to take on some of this risk, you would be

much more inclined to do it within a buyout vehicle because you

not only get to take longevity risk at perhaps a good price but you

also get control of assets that you can work over time.

Risk: Have you seen any trades of this kind in the market?

NT: Not directly in longevity derivatives, but the buyout market

has been very active. There has been a recent twist in the buyout

market, where Citi did a transaction with Thomson Regional

Newspaper (TRN). TRN effectively transferred the financial

responsibility for its pension fund to Citi. What is interesting is that

it allowed us to operate under pension fund regulations, not

insurance regulations, which may be more attractive in certain

circumstances. That is where a longevity market is needed. It’s not

as simple as hedging interest rate risk. It’s a trend that has started

and will continue to develop. We will see more and more transfers

because it’s a very efficient way to do this. It is natural for banks to

use an LDI approach after transfers. This is going to create a lot of

demand for longevity derivatives.

Risk: Are more institutions moving towards this structure? Were

there any regulatory hurdles that had to be crossed

NT: We are seeing more and more enquiries about this specific

transfer. A lot of people are interested in doing similar deals. In

terms of regulatory hurdles, there were a few things to check in

order to ensure that we could actually do this, as well as making

sure that members and trustees were kept in the loop.

Risk: Do we need to come up with more capital-efficient

structures before the market can gain traction?

TS: One of the ways I look at it is dividing up the markets and then

dividing up the risks within those markets. If you look at the US

pensions, there is the US defined-benefit pension market, which is

worth $2.3 trillion. Meanwhile, US variable annuities – most of which

have benefits that are hedgeable in this way – are worth $1.3 trillion.

It is a huge market place. The part of the market that has the most

capital markets solutions attached to it so far is the US variable

annuity market. There are banks that are stepping in and taking

longevity risk in those transactions, there are reinsurers that are

stepping in, there are those that are just taking the volatility aspects

of it. The insurance companies keep the longevity risk if they want to,

and there are plenty that don’t want to. But they don’t want the

volatility and the mark-to-market aspect of the liability, so they’ll ask

an investment bank to come in and take out that volatility. I may be

the most optimistic person at the table in seeing this market develop

and maybe that is because I see it developing so rapidly and so

much money being spent by insurers in the US to hedge these risks

today in a lot of different ways. They range from futures to cashflow-

structured over-the-counter hedges and everything in between.

YP: The trend is for risk to be transferred now that people are much

more aware of risk and the impact of volatility. That is going to be a

secular long-term trend – risk is going to be aggregated and

managed and it will be from different sources and disaggregated

into its different components or different risk aspects, some of

which can be hedged today and some of which will need to find a

market. It is a question of what price and how much capital you

can attract to take on that risk. That is something that everyone is

bullish about. What form that takes with regard to mortality and

longevity risk is hard to say at this point but it will happen, that risk

will get disaggregated from all the other risks that traditionally

come bundled together.

Risk: Is this going to be vanilla in terms of derivatives?

MP: I do not think that longevity derivatives will be merely designed

to address the pure longevity risk. They will also be tailored to meet

the demands of the new regulatory or accounting rules and their

underlying shift towards mark-to-market approaches. These typically

require giving a value to the mortality/longevity contingencies

embedded in products otherwise sensitive to market shifts.

EG: As with inflation it is an unbalanced market as end-users

ultimately get what they want. Banks end up with basis risk

between different inflation indexes or seasonality risk that they

cannot hedge but hope that they have priced correctly and will

balance out. It is hard to see mortality as being similar to that. The

market will need to find people who are willing to take on and

price basis risk in return for a risk premium and it is there that the

transfer of risk from end-users will take place.

BC: Swaps are the building blocks. The derivatisation of the risk

really happens when you have different interests and an active

market. The problem with the longevity market is that it is very

much a buy-and-hold market, so I don’t see volumes being very

large in terms of interbank dealing and, if they’re not very large, then

it is very difficult to go to the next level of derivatisation. It is going

to be difficult to have an interbank market on volatility, for example,

for a long time. That being said, because transactions are likely to be

bespoke, end-user structures could very well immediately be very

complex. Maybe one could have some optionality or something

exotic embedded in the first trade with a reinsurer on one side, a

pension on the other side and a bank providing the structuring and

keeping the unhedgeable correlation or volatility risk. But, in terms

of the development of the interbank market, it should be slow.

NT: The first step is to agree on a forward curve and we are not even

there yet. If there is going to be any complexity in the market, there is

more going to be because of different subindexes that people want

to trade rather than trading more complex structures that include

volatility features. The subindexes will be the drivers of the complexity.

Cyril Gonzalez

Head of index-linked derivatives

T. +44 (0)20 7200 7000

E. CGonzalez@tullettprebon.com

www.tullettprebon.com

Michaël Picot

1107_Risk_longevity.indd 49 30/10/07 11:18:40